Science on SciBench (Accuracy)
28.3Accuracy (SciBench)SC
Evaluation Results
| Method | Links | |
|---|---|---|
| SCBackbone=Qwen-2.5-72B-Instruct2025.05 | 28.3 | |
| AFlow-MathBackbone=Qwen-2.5-72B-Instruct2025.05 | 28.3 | |
| SingleBackbone=Qwen-2.5-72B-Instruct2025.05 | 28.1 | |
| MAS-GPTBackbone=Qwen-2.5-72B-Instruct2025.05 | 28.1 | |
| SCBackbone=Llama-3.3-70B-Instruct2025.05 | 27.9 | |
| DebateBackbone=Llama-3.3-70B-Instruct2025.05 | 27.9 | |
| MAS-GPTBackbone=Llama-3.3-70B-Instruct2025.05 | 26.9 | |
| DebateBackbone=Qwen-2.5-72B-Instruct2025.05 | 26.9 | |
| MADBackbone=Qwen-2.5-72B-Instruct2025.05 | 26.9 | |
| DyLANBackbone=Llama-3.3-70B-Instruct2025.05 | 26.7 | |
| AutoGenBackbone=Qwen-2.5-72B-Instruct2025.05 | 26.5 | |
| CoTBackbone=Qwen-2.5-72B-Instruct2025.05 | 26.3 | |
| AgentVerseBackbone=Qwen-2.5-72B-Instruct2025.05 | 25.9 | |
| AgentVerseBackbone=Llama-3.3-70B-Instruct2025.05 | 25.7 | |
| SingleBackbone=Llama-3.3-70B-Instruct2025.05 | 25.5 | |
| AFlow-MathBackbone=Llama-3.3-70B-Instruct2025.05 | 25.3 | |
| MacNetBackbone=Qwen-2.5-72B-Instruct2025.05 | 25.1 | |
| CoTBackbone=Llama-3.3-70B-Instruct2025.05 | 24.7 | |
| DyLANBackbone=Qwen-2.5-72B-Instruct2025.05 | 24.7 | |
| MAVBackbone=Llama-3.3-70B-Instruct2025.05 | 24.3 | |
| MADBackbone=Llama-3.3-70B-Instruct2025.05 | 23.7 | |
| MacNetBackbone=Llama-3.3-70B-Instruct2025.05 | 23.5 | |
| AutoGenBackbone=Llama-3.3-70B-Instruct2025.05 | 21.9 | |
| MAVBackbone=Qwen-2.5-72B-Instruct2025.05 | 14.5 |